Beyond Cell Types: How SeekOne™ DD scFAST‑seq Supports Tracing the Origins and Immune Ecosystem of Fusion‑Driven AML

Beyond Cell Types: How SeekOne™ DD scFAST‑seq Supports Tracing the Origins and Immune Ecosystem of Fusion‑Driven AML

A PNAS study leverages fusion‑aware single‑cell transcriptomics to connect leukemia genotype, differentiation hierarchy, immune remodeling, and clinical risk.

Fusion‑driven acute myeloid leukemia (AML) is usually viewed through malignant blasts. But do fusion‑bearing stem or progenitor cells also seed apparently mature immune and erythroid lineages?

A 2026 PNAS study addressed this question in 24 newly diagnosed patients with RUNX1::RUNX1T1, PML::RARA, or CBFB::MYH11 AML. By combining single‑cell transcriptomics, fusion detection, DNA‑FISH, and targeted mutation sequencing, the investigators traced leukemia‑associated fusions across the hematopoietic hierarchy.

The scientific challenge: connecting genotype, cell identity, and cell state

Standard 3′/5′ single‑cell RNA‑seq defines cell types and states, but may miss internal fusion junctions, mutations, or splice events. Fusion‑driven AML therefore requires a cell‑resolved link between molecular genotype and transcriptional phenotype.

The key questions are:

• Which lineages carry the leukemia‑defining fusion?
• How does fusion status reshape differentiation and immune state?
• Can persistent clones or communication patterns inform relapse and risk?

Answering them requires transcript‑wide expression and targeted fusion evidence at single‑cell resolution.

Why SeekOne™ DD scFAST‑seq was selected in this study

A dual‑platform strategy for robust fusion tracing

The study analyzed bone marrow from 24 patients: eight with RUNX1::RUNX1T1, seven with PML::RARA, and nine with CBFB::MYH11 AML. Two complementary workflows were used: 10x Genomics plus targeted Oxford Nanopore sequencing, and SeekOne™ DD scFAST‑seq plus Panel‑seq.

SeekOne™ DD scFAST‑seq distributes short‑read fragments across transcripts; the resulting whole‑transcriptome cDNA can also feed a targeted panel for leukemia‑associated genes. This connects cell identity and state with fusion evidence.

After quality control, 234,018 transcriptomes remained, including 46,737 cells with directly detected fusion transcription factor expression. Negative controls helped test signal specificity.

Finding 1: leukemia‑associated fusions cross lineage boundaries

The atlas detected fusion transcripts in blasts and progenitors, but also in monocytes, dendritic cells, T/NK cells, B cells, and erythroid cells. To address transcript dropout, direct detections were combined with subtype‑specific expression signatures to define putative fTF‑positive cells.

DNA‑FISH confirmed genomic fusions in sorted T cells and, where tested, erythroid cells. Secondary‑mutation profiling supported a stepwise model proposed by the authors: the founding fusion may arise in HSCs/MPPs, while later lineage‑restricted alterations drive overt AML.

Finding 2: each AML fusion creates a distinct differentiation hierarchy

All three subtypes showed an upstream HSC/MPP‑like state but developed distinct downstream architectures.

• PML::RARA AML expanded GMP‑like cells
• RUNX1::RUNX1T1 enriched a specific blast population
• CBFB::MYH11 showed a separate blast state with more dendritic and monocytic cells

The atlas therefore resolved fusion‑associated differentiation blocks and subtype‑specific trajectories observed by the research team.

Finding 3: persistent fusion‑associated cells distinguish relapse from durable remission

Five RUNX1::RUNX1T1 patients were followed through diagnosis, remission, or relapse. Fusion‑associated T/NK, B, and erythroid populations persisted in patients who relapsed, but markedly declined during durable remission, falling below one per thousand cells in lymphoid compartments.

These observations suggest an upstream fusion‑bearing reservoir and make mature fusion‑positive cells a testable candidate for residual‑disease and preleukemic‑hematopoiesis studies. Their functional role remains unresolved.

Finding 4: single‑cell discovery can support clinically relevant models

Ligand‑receptor analysis identified subtype‑specific communication networks among blasts, putative fusion‑positive immune cells, and fusion‑negative microenvironmental cells. In RUNX1::RUNX1T1 AML, five ligand‑receptor pairs were combined into a risk score.

The score stratified event‑free and overall survival in non‑transplanted patients and was supported in independent datasets. It remained an independent predictor of event‑free survival in the tested cohort after clinical adjustment.

From a fusion event to a disease ecosystem

The evidence supports a model proposed by the authors in which a founding fusion arises in an HSC/MPP, crosses hematopoietic lineages, and is followed by secondary alterations that create overt AML, subtype‑specific differentiation hierarchies, and potentially distinct immune networks.

SeekOne™ DD scFAST‑seq provides a short‑read workflow that enables researchers to interrogate this biology by connecting transcript‑wide information with single‑cell identity, supporting research into fusion‑driven cancer, clonal evolution, residual disease, immune dysfunction, and treatment resistance.